英语语音合理性优化识别建模仿真研究

发布时间：2018-01-22 04:06

本文关键词： 英语发音音准合理性识别　出处：《计算机仿真》2017年02期 　论文类型：期刊论文

【摘要】：对英语语音合理性进行准确识别,可以提高英语发音的准确性。在英语发音过程中,需要对在应力影响下出现变异语音进行过滤,但是传统的识别模型进行音准识别时,主要通过发音音准特征进行识别,无法消除应力影响下的变异语音,导致建模误差大的问题。提出基于改进粒子滤波的英语语音合理性优化识别建模方法。采集不同人的英语发音和其唇动信息,依据信息论中的最大互信息准则与盲源分离中的高阶统计量准则,获取英语发音音准合理性目标函数,并融合于动态时间规正原理,构造英语发音音准差别子空间,消除应力影响下的变异语音,对英语发音音准的主体部分进行识别,并将英语发音语音特征矢量进行长度对齐,组建英语语音合理性优化识别模型。仿真结果表明,利用优化模型进行英语语音合理性识别时识别精度较高,为英美发音差异教学提供了一定的科学依据。
[Abstract]:Accurate recognition of the rationality of English pronunciation can improve the accuracy of English pronunciation. In the process of English pronunciation, it is necessary to filter the variant speech under the influence of stress. However, when the traditional recognition model is used to recognize the sound level, it can not eliminate the influence of the stress on the variant speech. This paper proposes an improved particle filter (Particle filter) based method to optimize the recognition and modeling of English speech, and collects the information of different people's English pronunciation and lip movement. According to the maximum mutual information criterion in information theory and the high order statistics criterion in blind source separation, the objective function of English pronunciation rationality is obtained and integrated into the dynamic time normalization principle. In this paper, we construct the subspace of English pronunciation accuracy difference, eliminate the variation speech under the influence of stress, recognize the main part of English pronunciation accuracy, and align the length of the English pronunciation speech feature vector. The simulation results show that the recognition accuracy of the optimization model is high, which provides a scientific basis for the differential teaching of English pronunciation.
【作者单位】：黄河科技学院外语学院;
【分类号】：TN912.3
【正文快照】： 1引_言英语教师在语音教学过程中对英语发音音准合理性进行讲解和示范,促使学生了解英语发音的多样性,可以提高学生英语语言听说能力[1_3]。但目前国内的高校中的语音教学难以提升学生的英语发音质量[44]。在此背景下,有效地进行英语语音合理性识别,可以解决上述问题,具有重

【相似文献】